THE APPLICATION OF STANDARD GENERALIZED AUTOREGRESSIVE CONDITIONAL HETEROSCEDASTICITY (SGARCH) MODEL IN FORECASTING THE STOCK PRICE OF BARITO PACIFIC

Abstract

Stock potentially yields higher returns than other investments, but is riskier due to volatile prices. To minimize the risk of loss, investors can forecast the stock price to help in deciding whether to buy, sell, or hold the stock. Several methods are available for forecasting the stock price such as ARIMA, ARCH, and SGARCH. ARIMA model works best for series with a constant variance of error. However, almost all stock price series have a non-constant variance of error, known as heteroscedasticity, as such ARIMA isn’t suited for modeling the stock price. In contrast, the SGARCH model can handle series with heteroscedasticity. This makes it better suited for modeling stock prices as they have similar characteristics. PT Barito Pacific (BRPT) is a publicly traded firm that works mainly in petrochemical and geothermal energy. BRPT’s net profit increased in 2023 by 243% and the demand for geothermal energy is expected to increase due to the government's renewable energy transition project. Therefore, this study forecasts the BRPT’s stock price using the SGARCH model with R Studio. The stock price used ranges from October 1st, 2018 to August 16th, 2023 gotten from the Yahoo Finance Website. Based on the least AIC, this study found that ARMA(6,2)-SGARCH(1,1) is the best model for forecasting the stock price. This model gives a very accurate prediction of the stock price from April 1st, 2023 – April 19th, 2023 with a mean absolute error of 78.11, root mean square error of 89.51, and mean absolute percentage error of 9.81%.